Variable Primary Chilled Water Systems Part 1: The Relationship Between Chiller Tons, Flow & Delta-T
/By Chad Edmondson
What is a variable primary chilled water system? How does it differ operationally from a primary/secondary system? And why might you choose one chilled water design approach over the other?
In this series on Variable-Primary Chilled Water Systems, we’ll be answering all of these questions while exploring other related topics such as staging, low delta-T syndrome, etc. But before we jump into any of that, we need to answer one very fundamental question just in case anyone reading this is at all new to chilled water design: What is a chiller (or air conditioning) ton?
At the risk of explaining something that no doubt most of you already know, a ton is a measurement for how much heat (BTUs) a chiller or other air conditioning unit can remove from a space in one hour. It takes one ton of air conditioning to remove 12,000 BTUs in an hour. Therefore, a five ton chiller can remove 60,000 BTUs (5 x 12,000) in an hour.
In the HVAC industry, we refer to tons and tonnage endlessly – perhaps so much so that the meaning no longer even registers. But we promise if you keep this one fundamental definition in mind, everything else we talk about in this series will be far easier to digest and apply. Because that’s what it is all about, right? We want to design a system that allows us to meet that building cooling load.
Tonnage, Flow & Delta T
In any system, BTUH = 500 x GPM x ΔT. (Note: 500 is a constant that is based on the weight of one gallon of water (8.33 pounds) times 60 minutes in an hour. So, 8.33 pounds x 60 minutes = 500). To convert this to chiller tons, we simply divide by 12,000 since there are 12,000 BTUs in one cooling ton:
It follows then that if we have a 1000 GPM chilled water system with a 44°F supply temperature from the chiller and 56°F return temperature to the chiller, giving us a 12 degree ΔT, we know we need 500 tons of cooling capacity:
Under these conditions, our 500 ton chiller would be fully loaded and it would be delivering exactly 500 tons of cooling. But what happens to the chiller output (and by output, remember we mean its ability to remove BTUs) if we have only a 10 degree ΔT? As always, the laws of math and physics prevail and as you can see by the calculation below, we are no longer able to get 500 tons of cooling capacity from our chiller. In other words the chiller is only 83% loaded.
This is not a problem as long as our building load also happens to be at 83%. However, if we are fully loaded on the building side, a reduced ΔT means that (1) something is amiss somewhere out in the system and (2) we cannot deliver full cooling capacity under the current conditions.
But what about our other variable in the above equation – GPM? If we had the option of varying flow through the chiller, we might be able to mitigate this problem by temporarily increasing the flow. In this case, if we increase our chiller flow to 1200 GPM, we can reestablish the 500 tons that we need:
Is there a penalty for increasing the flow? Of course. In this case, increasing the flow to 1200 GPM will increase the chiller pump break horsepower through the chiller by 70%. (See our blog on Pump Affinity Laws). That may or may not be a solution for the given situation, but you begin to see how it might be beneficial to vary flow through the chillers. More importantly you begin to understand the relationship between flow, delta-T and tonnage.
Next up, we’ll explore the problems associated with both low and high return water temperatures on primary secondary systems and what solutions may be available on these systems where varying flow through the chillers is not an option.
